building techniques and materials, Roman
building techniques and materials, Roman
- Roger B. Ulrich
The inherent strengths, weaknesses, and availability of diverse Roman building materials governed the techniques used in construction and greatly influenced the final appearance of Roman architecture. Trace archaeological evidence exists of buildings and burials in Rome from the Italian Bronze Age (second millennium bce) or earlier, and substantial physical remains, in the form of Iron-Age huts and grave goods, roughly correspond to the Romans’ own belief of the foundation date of their city (traditionally 753 bce). Rome’s earliest builders sourced materials obtainable from the immediate environment and transformed them using practical knowledge. Within the span of a couple centuries, architectural design, implementation, and decoration reflect a broad interaction between Roman builders and their counterparts in the regions around central Italy (particularly Etruria to the north and Campania to the south) and also the wider Mediterranean world, particularly those areas where Greeks traditionally lived or had placed colonies. While southern Italy and Sicily represent the closest areas for the transmission of Greek ideas, Greek building practices on the Greek mainland and in Asia Minor also influenced Roman projects from the Archaic period onwards. As Rome grew wealthier and expanded abroad, patrons and builders imported marble to the capital from the Aegean, well before the discovery of more local, Italian sources. The importation of exotic stones grew exponentially over the period of the late Republic and the first two centuries of empire. The coloured marbles that embellished the buildings of Rome served as physical testimony to Rome’s control over the eastern Mediterranean. Nothing, however, was as transformative as the adoption of concrete in the late 3rd century bce, the mass production of fired brick, and the ensuing experimentation that resulted in the vaulted structures that have become the hallmark of Roman architecture.
- Roman Material Culture
Topography and Natural Resources
Long before Roman emperors imported coloured stones from different regions of Italy, from North Africa, and from the lands of the eastern Mediterranean and Egypt, Rome and its environs were perfectly self-sufficient in terms of durable and accessible building materials. Rome itself rises over deposits of volcanic tuff (see Rome, topography). The strata of soft tuffs lie close to the surface and are relatively easy to quarry. To the east of the city extend thick deposits of harder travertine. Volcanic basalt (silex or selce), hard and durable, paved Roman roadbeds from the 3rd century bce. Dense forests of soft and hard woods, covering the high hills and mountains to the north and east of Rome, furnished the long, straight beams used for framing and roofing buildings. When local supplies were exhausted, the Tiber River facilitated the transport of heavy stone and log-length timbers from the interior of Italy to the docks of Rome. An important tributary just to the north of the city, the Anio River, was another important means of accessing prime building materials from the interior. Significant limestone deposits to the north and east provided stone for the capital’s first colonies in Latium and Etruria and later supplied the essential ingredient, calcium chloride, for the mortars used in concrete construction from the mid- to late-Republic onwards.1 In addition to the variety of tuffs and lava stone, volcanic activity in the area had also produced copious amounts of an “ash” that when mixed with lime created a robust mortar that had the seemingly magical property of being able to harden underwater. Deposits of this substance lay only a few kilometres from the city centre of Rome. Beds of clay, including significant deposits right in the middle of the city, provided the raw material for fired roof tiles, architectural revetments, and bricks. A short distance up the coast from Rome a “metal-bearing” range of mountains (Catena Metallifera) initially yielded significant amounts of iron ore; blacksmiths forged the smelted iron into millions of fasteners, such as iron clamps and tie bars that strengthened both stone and concrete construction.
Beginnings: Wood, Clay, and Unworked Stone—Ephemeral Forms and Materials
Vitruvius considered building with local materials one of mankind’s first civilizing activities, taking place immediately after the discovery of fire and the development of language (De arch. 2.1.1–2). In his view, the first construction materials were those most readily at hand: leaves, sticks, and mud or clay. He imagined primeval builders observing and imitating the practices of animals. Thus the intricate woven nests of swallows offered a template for the invention of wattle and daub, a construction technique characterized by the packing of mud or plaster upon a tightly woven panel of twigs (De arch. 2.1.2). These ur-craftsmen, he posited, possessed a nature both imitative and teachable (imitabili docilique natura), qualities not lacking among Vitruvius and his peers, who studied and learned from the architectural achievements of their predecessors, and who were not adverse to imitating and repurposing the models they studied, whether Italian, Greek, Egyptian, or Punic.
Unworked wood (“roundwood”), clay, and stone were the primal materials for building and never fell out of use. Vertical forked sticks (furcae) supporting horizontal poles of saplings or trimmed limbs defined the fundamental elements of post-and-lintel architecture that would later be realized in shaped dry stone and then concrete masonry. The funerary “hut” urns of central Italy from the Italian Iron Age (10th–8th centuries bce) and contemporaneous post-hole remains in Rome and Latium of actual dwellings attest to this type of roundwood post-and-beam framing filled with panels of wattle and daub. From here it was a relatively short step to framing walls with a squared grid of vertical and horizontal squared timbers and filling the gaps with packed stones and clay (7th century bce). Much later (from the 2nd century bce) the substitution of concrete for the infill would characterize the widespread use of opus craticium (see “Roman Concrete” section).
Depending on soil density and composition, walls of earth must have been one of the earliest forms of materials used by early Italian builders. Vitruvius (De arch. 2.1.3) believed that moist “clods” (luteae glaebae) of soil strengthened by wooden timbers made suitable walls, a practice that continued well into the Imperial period in various forms, such as camp walls built by legionaries on frontiers that were constructed of sod and bound by cross timbers and “corduroy” walkways. Certainly the method he describes was still being used in his day in Gaul. Walls of unfired mud brick have been found in many parts of the Roman world, from all periods, and rammed earth (pisé, or paries formaceus) construction is described by Pliny as characteristic of North Africa and Spain (Plin. HN 35.169).
The development of iron tools, particularly those with sharp cutting edges—including axes, adzes, picks, saws, drill bits, and chisels—over the course of the Archaic period greatly facilitated the extraction and processing of raw materials and the speed of execution and scale of architectural projects.
Raw Materials and Techniques of Construction
Wood (see carpentry, Roman) was an essential component of Roman construction in all periods and in most parts of the Roman world. Wood played a key structural role due to its strength both in a state of compression and in tension. The former made wood suitable for vertical supports. The latter made it indispensable for horizontal beams, especially for the tie beams of timber trusses employed to span broad spaces up to one hundred Roman feet across.
Romans knew that some of their neighbours in heavily wooded areas built structures entirely out of wood. The easiest and the most wasteful way to build entirely of wood was by using notched logs, an exotic practice described by Vitruvius (De arch. 2.1.4). Planks were easier to handle and served as excellent sheathing for walls and roofs, as well as for interior finish work including door and window frames, the leaves of doors, and window shutters. Planks required much more labour as they were generally cut by hand in saw pits from blank stock of squared timbers, although in some extraordinary cases water-powered sawmills have been documented in the provinces (e.g., as depicted on a 3rd-century ce sarcophagus from Hieropolis in Asia Minor).
Wood played a role in all phases of construction and in all parts of the building, even for foundations. Horizontal “sleeper” beams were placed in shallow trenches and supported the upright timbers used for walls; the practice was widespread in northern Europe and Britain. The ends of wooden piles could be sharpened and driven into the ground or into riverbeds to support buildings and bridges. For the latter, clusters of wooden beams were used to support superstructures even when the piers emerging from the water were built of stone (e.g., at Roman Trier, 1st century ce). Rome’s earliest permanent bridge across the Tiber was made entirely of wood (the Pons Sublicius, the “pile bridge”). Wooden piles suspended buildings above the ground in areas prone to flooding. Strabo mentions such construction for the area of Ravenna (Geog. 5.1.7), and planked structures on piles represented on Trajan’s Column (113 ce) are recognized by scholars as a form of Dacian architecture. Porticos and temples dating to the Archaic period employed columns of wood. These could be partially sheathed with terracotta, as has been documented from a temple of the Omobono complex in Rome. Buried wooden posts strengthened roadbeds crossing marshy areas. At Rome’s port town of Ostia, excavators discovered wooden posts and crossbeams used below grade to stabilize the superimposed roadbed. Romans were well aware that some species resisted rot better than others when exposed to damp conditions. They also treated wood with various concoctions or charred it to help with rot resistance. When submerged below the level of ground water, wood is very resistant to decomposition; wooden pipes were sometimes substituted for lead to channel water to and from buildings. Later, when concrete was favoured for the foundations of large structures during the Imperial period, wood played an important role as the rigid form (“shuttering”) that held the concrete in place until it could cure.
Wood was used for walls by itself or with other materials (fig. 1).
Opus craticium, a method characterized by a wooden framework with an infill of concrete, was widely employed for upper storeys and interior walls; good examples can still be seen at Herculaneum, where the exposed wooden framing has carbonized from the hot pyroclastic flow from Vesuvius and is thus preserved.
Horizontal wooden architraves spanned openings in walls and the intercolumniations of porticoes. Romans learned that composite beams (trabes conpactiles, Vitr. De arch. 5.1.8) comprising a “sandwich” of planks offered great strength when single beams were not available.
While Roman builders could use vaults (see “Vaulting in Stone, Brick, and Concrete” section) to support upper floors—a practice of increasing use in the Imperial period—wood served as a readily available and cheaper alternative. The Latin term contignatio refers to a practice of wooden flooring that was most commonly a composite of wooden framing that supported a thick layer of rubble (rudus) finished with a pavement of tile, crushed terracotta mixed with mortar (cocciopesto or opus signinum), herringbone brick (opus spicatum), or mosaic. Such floors comprised a set of heavy support joists of squared or round timbers embedded into the side walls, one or two layers of floorboards attached with iron nails, and top layers of masonry (Vitr. De arch. 7.1.2). By the 2nd century ce it was common to use only a few heavy joists that in turn supported a second deck of cross timbers capped by floorboards.
Roofing represents the most prevalent use of wooden framing in construction. The relatively light weight of wood and its load-bearing capacity were ideal characteristics for covering buildings (fig. 1a).
Houses and temples of the Archaic period were generally covered with beams employed in what is known as a “prop-and-lintel” system, where heavy purlins, ridgepoles, and the rafters they carried were supported by vertical “props” placed directly over interior walls and columns.
Scholars have argued that the triangular tie-beam timber truss was used as early as the Archaic period. Its full potential (and perhaps its actual introduction) is seen with the construction of the great civic basilican halls that first appeared in the 2nd century bce. The triangular truss was capable of clearing spans of up to one hundred Roman feet, surpassing even the greatest of the cross vaults (see “Vaulting in Stone, Brick, and Concrete” section). Timber trusses covered basilicas, Imperial audience halls (including the great halls of Domitian’s palace), senate houses (including the Curia of Rome), and covered theatres (like the Odeon of Pompeii). The inwardly “impluviate” sloping roofs of Roman atrium houses were all framed with wood, and wooden beams supported interior and exterior balconies and the shed-type roofs that projected over doorways and sidewalks.
Wooden spans bridged rivers during both the Republican and Imperial periods. Julius Caesar’s famous description of the bridge he built over the Rhine took the form of a series of trestles that supported horizontal timbers; the passage offers important Latin vocabulary for such structural members (Caes. BGall. 4.17). Other bridges may have employed cantilevers. Apollodorus of Damascus is credited with designing the ambitious Trajanic-period bridge over the Danube; it employed wooden segmental arches paired with stone piers. At least one depiction (on Trajan’s Column) and literary description (Dio Cass. 68.13.1) attest to the achievement.
Most doors, door and window frames, and window shutters were made of wood. Double-leaved wooden doors swung on bronze pivots (cardines) set into the floor and lintel; in lighter applications hinges were made of bronze and bone. Romans developed strong glues from fish or animal cartilage to join planks edge to edge. The ubiquitous Roman tabernae of cities and towns used a series of floor-to-lintel wooden planks to shutter their shops at night.
Wood played an important role even when not destined to form a permanent component of a finished structure. Its use for the framing of concrete foundations has been mentioned. Large quantities of rough beams or poles formed the scaffolding required to construct walls of masonry. Surviving putlog holes in walls of brick attest to its use. With few exceptions (e.g., vaulting tubes), Roman forms and methods of vaulting required preliminary centring work of suitable beams and planks (fig. 2).
No Roman description of centring survives; even the Latin term for it is uncertain, but its employment is manifest everywhere, from the imprints of the boards used still visible on the undersides (intradoses) of concrete vaults to the projecting stone blocks left on aqueduct projects (e.g., Pont du Gard, 1st century ce) and other buildings that supported the centring while the vault was under construction.
Finally, the machines that lifted heavy materials into place during construction were fashioned primarily of wood. Cranes large and small, including the great wheeled example depicted on the relief from the Tomb of the Haterii in Rome (late 1st century ce), were essential equipment at Roman building sites, including all manner of wooden windlasses used to jostle heavy stones into position.
Terracotta and Brick
The practice of using fired clay for sheathing roofs and exposed wooden beams began in Rome in the second half of the 7th century bce. Raw materials, organized labour, and scaled-up technology necessary for manufacturing the large numbers of tiles required to cover a roof appear at just about the same time that Roman houses and public buildings expanded from the hills of the future city to the area of the forum valley. A large dwelling that stood in the “Sepulcretum” area of the future Roman forum sometime after 650 bce offers one of the earliest examples. By the first quarter of the sixth century bce, roof tiles and terracotta revetments protected the forum’s earliest buildings serving a public function; the Regia is a prime example. Excavated terracotta fragments, including antefixes and friezes, exhibit orientalizing motifs like gorgon heads, minotaurs, and animal processions.2 Terracotta roofing tiles remained in use throughout the Roman period, while terracotta revetments faded in importance at the beginning of the Imperial period as stone entablatures and concrete vaulting replaced wooden beams.
Good beds of clay suitable for roofing tiles and architectural terracottas have been found in the area of the Velabrum in Rome; the use of local clay suggests local production. Terracotta revetment plaques were nailed to the wooden beams of the superstructure. To eliminate cracking, holes for the nails were set in the clay before firing.
At the sanctuary site of Sant’Omobono, fired clay collars capped wooden column shafts of a shrine dated to the 6th century bce (fig. 3). Large-scale figures in terracotta decorated roofs and pediments.
Fired clay would continue to play a leading role in Roman construction in central Italy. Even provincial cities of the Imperial period that had neither the tradition of working with clay nor good sources of the raw material show evidence of Italic-style brickwork in some applications like apses and vaulting (e.g., the Severan basilica at Lepcis Magna, early 3rd century ce).
Fired clay bricks that formed the facing for concrete walls (see “Roman concrete” section) and vault construction developed directly from the production of roof tiles; tiles with their flanges removed served as facing for the earliest brick walls in the 1st century bce. Mass production of fired brick from the early Imperial period onwards resulted in more and more construction with brick as the primary facing material. The practice of stamping bricks before firing (usually at a temperature of c. 800° C) with an identifying mark has proved to be an indispensable aid for dating construction.3 Walls of solid brickwork (e.g., the temple of Serapis at Pergamon, 2nd century ce) are less common and associated with provincial architecture in Asia Minor.
A standard bessalis (two-thirds of a Roman foot square, about 20 centimetres) or sesquipedalis (1.5 Roman feet square, about 45 centimetres) were cut on the diagonal to furnish the characteristic triangular bricks used for wall facings.
The largest standard size, the “two-footer” (bipedalis) paved utilitarian floors (fig. 4), defined string and leveling courses in walls, spanned the small piers (pilae) of hypocaust heating systems, and strengthened vaults.
Bricks of different colours and shapes enhanced some buildings or created inexpensive shortcuts for creating architectural elements like columns (fig. 5).
A layer of stucco over a brick column provided the illusion of more expensive solid marble.
At Rome and the nearby port city of Ostia, molded bricks replicated column bases and entablature moldings, particularly in utilitarian buildings like warehouses and for private tombs.
Small special-purpose bricks were laid on edge in a herringbone pattern (opus spicatum; see fig. 13) for a durable, attractive, and inexpensive floor. Crushed tile or brick mixed with mortar (cocciopesto, also described as opus signinum) resulted in a tough waterproof coating for floors, cisterns, and flat roofs.
Special-use fired bricks with customized forms appeared in applications where builders sought resistance to heat and moisture or opted for ready-made materials to ease construction (fig. 6).
Stacked terracotta tiles or tubular miniature “columns” supported hypocaust floor systems used in baths, palaces, and elite domestic spaces. “Nipple tiles” (tegulae mammatae) allowed hot air to circulate within the intermural space of heated rooms. Rectangular flue tiles built within walls achieved a similar purpose.
Cylinders of fired clay encased within a wall provided channels for the drainage of wastewater, just as terracotta pipelines coursed beneath city streets.
Fired clay played an increasingly important role in Roman vaulting. Roman walls of faced concrete opus caementicium commonly incorporated brick-relieving arches over areas of weakness, like doors, windows, or deep interior niches. Various forms of brick ribbing appear in Roman concrete vaults from the 1st century ce; by the 4th century, brick ribbing was integral to concrete vaulting in Rome.4 Interlocking tubes of terracotta arranged in closely spaced rows made it possible to build vaults without wooden centring; the practice is first documented from the 3rd century bce in Sicily and was particularly popular in North Africa.5 For the baths of Roman Britain, builders even experimented with barrel vaults fashioned from voussoirs of hollow fired brick.6 Such use would have lightened the vault, a clear desire of high-late Imperial builders who often incorporated empty ceramic amphorae into vaulting in an effort to lessen the overall mass of the construction.7 Amphorae have also been found buried under walls to stabilize loose or wet substrates and to improve drainage; the practice has been documented in the area of the Po watershed. Recent excavations in the Monte Testaccio area of Rome have revealed that tightly-packed “recycled” amphorae placed in superimposed rows formed the walls of some utilitarian buildings in the port area.
The cutting away and shaping of bedrock to form postholes for interior wooden posts and the low sills of interior walls, or the simple cylindrical or rectangular chambers for Iron-Age graves, represent the earliest form of Roman architectural stonework. Unworked stone, some merely in pebble form, lined tombs and served as simple foundations or pavements.
To the north of Rome, the walls of funerary tumuli already employed locally quarried sandstone in the orientalizing period (e.g., at Etruscan Populonia in the 7th century bce). Sixth-century Rome witnessed rapid expansion of squared-stone construction (opus quadratum) with local tuffs. Ashlars of tuff cut to uniform dimensions were employed extensively well into the Imperial period for foundations, fortifications, and temple podia—virtually any application where strength and durability were desired. It has been argued that the first attempts at stone vaulting (see “Vaulting in Stone, Brick, and Concrete” section) can be dated to the 6th century bce.
The use of stone for Classical entablatures appeared by the end of the 2nd century. For most of the Archaic period, horizontal wooden beams, often revetted with terracotta, served in this role. Wooden architraves are superior to stone in terms of tensile strength, thus early porticoes and colonnaded porches exhibit broad intercolumniations. Where wood was not practical for spanning (as in subterranean or other damp conditions), corbelling or paired raking slabs of stone were employed for short spans. The tendency of stone to crack under tension resulted in shorter intercolumniations for temples and porticoes built entirely with stone. Wood and stone could be paired to span broader intercolumniations. At the south portico of the forum of Pompeii (1st century bce) wooden architraves once supported upper frieze and cornice courses of tuff stone. The so-called flat (or lintel) arch offered another solution for achieving wide horizontal spans in stone; voussoirs of stone in horizontal courses stayed in place through gravitational force and lateral compression (fig. 7).
The new portico still under construction in Pompeii’s forum in 79 ce offers an example in a local marble, as does the finely cut “keyed” stonework of Diocletian’s palace at Split built nearly 250 years later.
Horizontal iron bars integrated with stone (or brick) architraves also served to relieve stress and permit broader intercolumniations than were possible with monolithic stone blocks (e.g., the brick and stone composite lintels of the hall of the Doric pilasters, Hadrian’s Villa, 2nd century ce).
Summary of the Main Categories and Properties of Building Stone
Tuff: Tuff (Italian “tufo,” in many English textbooks “tufa”) deposits in central Italy are extensive and easily accessed; there are several good sources in and around the city of Rome. Cappellaccio tuff, a friable type of inferior quality, is Rome’s most local and earliest-used type. Most tuffs are relatively easy to quarry and can be cut into uniform blocks. Important deposits adjacent to the banks of rivers (e.g., “Anio” tuff) facilitated the transport of blocks downriver, especially to areas where there was no suitable local building stone (the first walls of Rome’s port town of Ostia, of Fidene tuff, are a good example). Tuff is very strong in compression, making it suitable for pavements, walls, piers, voussoir arches, and columns.
The podium of Rome’s immense Archaic temple to Jupiter Optimus Maximus (late 6th century bce) was of ashlars of cappellaccio stacked without mortar or clamps (fig. 8).
Blocks of “grotta oscura,” a yellow tuff (territory of Veii), make up the Republican “Servian” wall circuit (4th century bce); the precinct wall of the Forum of Augustus (dedicated 2 bce) is made from “peperino” tuff (Alban Hills). Tuff fractures easily under tension, so it is a poor choice for lintels and architraves. This inherent flaw encouraged builders to look for stronger stone (the travertine quarries near Rome provided one solution). The same properties may have promoted the development and rapid adoption of stone arches. Voussoirs of tuff, subjective to compressive force, were durable and capable of supporting great weight.
Travertine: Having discovered the vast travertine deposits to the east of the city along the Via Tiburtina, Roman builders used blocks of the hard, porous white stone in concert with softer tuff; the former was used sparingly at first (2nd century bce), placed at points of stress or where its tensile strength was superior. By the late Republic, travertine architraves were paired with upper entablatures and columns of tuff (Temple of Janus, Forum Holitorium, restored 1st century bce), used as keystones in tuff arches, and eventually used in foundations under points of concentrated weight (fig. 9).
Coatings of stucco on mixed tuff and travertine construction presented surfaces uniform in texture and color. Perimeter walls of tuff and travertine formed masonry “boxes” filled with opus caementicium to serve as temple podia. By the late 1st century bce, the supply of travertine was copious enough that the hard stone could be used for entire buildings (e.g., the facade of the Theatre of Marcellus, 13 bce). Travertine was also used as ribbing to reinforce barrel vaults: applications include the vaults of the tabernae in the Sanctuary of Hercules Victor (early 1st century bce).
Limestone: In areas to the north, east, and south of Rome, limestone was abundant for building.
Stones quarried from hillsides were dressed and assembled without being squared to form polygonal walls (opus siliceum) for colonial fortifications and temple podia (e.g., Cosa, 3rd century bce) (fig. 10).
Limestone could be cut into blocks of varying size and could be employed as a facing material for concrete walls, especially opus incertum and opus reticulatum (see below). Limestone also yielded lime used for concrete construction.
Marbles and granite: Q. Caecilius Metellus is credited with building the first marble temple in Rome in 146 bce (perhaps that of Jupiter Stator in the Porticus Metelli; Vell. Pat.1.11.5). Here and elsewhere (e.g., the late 2nd-century Temple of Hercules of the Forum Boarium), quarries in Greece supplied the stone. The Italian quarries at Luna represented the most important Italian source of white marble beginning in the mid-1st century bce and were in full production by the reign of Augustus. The emperor’s boast of recreating Rome as a city of marble was possible due to the high-quality stone shipped down the west coast from Luna. Rome’s control of the Mediterranean basin by the end of the 1st century bce provided access to fine marbles and other exotic stones from the entire region, with the most important centres in North Africa, Greece, and Asia Minor. Provincial cities and their local patrons throughout the empire benefitted from access to colourful marbles. Mainstays of the extensive marble trade included yellow-veined giallo antico from the Chemtou quarries of present-day Tunisia, green cipollino from Euboea, purple striated stone (pavonazzetto) from Asia Minor, and hard red porphyry from the quarries of Mons Porphyrites in Egypt. Egypt also supplied the best varieties of red and gray granite. Barges floated the stones to Alexandria for sea transport to Rome. Ships offloaded marbles at Ostia where the Tiber provided passage to its final destination. Columns and floor and wall revetments of imported marbles graced both public and private buildings. Translucent stones (like alabaster) could serve as windowpanes.
Basalt, lava, scoria: The extinct and active volcanoes of Italy produced durable stone suitable for road surfaces and foundations. Immense quantities of heavy, fine-grained basalt (selce) formed the distinctive paving of the major roads leading from the capital. Roman towns around Vesuvius employed a gray lava stone for the same purpose. Lighter scoriae, volcanic rocks filled with air spaces (vesicles) left from volcanic gases, were used as aggregate (caementa) with mortar to make low density concrete well suited for upper walls and vaults.
The date of the earliest walls in Rome with a structural core of lime-based mortar mixed with rubble (“concrete,” opus caementicium) is a topic of debate and revision; in Rome and Campania the medium was certainly in wide use by the mid-2nd century bce.8 While both Greeks and Romans made use of lime mortars as effective bonding agents, Romans discovered the efficacy of volcanic “powder” (pulvis puteolanus) mixed with slaked lime to produce a mortar with superior compression strength and durability. There were no good sources of lime in Rome or its immediate vicinity. This as well as the process needed to extract lime from the stone explain why concrete construction represents a later development in building technology. The limestone hills to the east of the city—the Monti Cornicolani, Tiburtini, and Prenestini—offered good sources for the calcium carbonate in limestone (the compound is present in travertine and marble as well). Labourers filled cylindrical kilns with chunks of limestone. After a period of slow firing the powdered quicklime was ready for slaking (mixing with water).
Slaked lime could be used immediately or dried and stored. Adding water, then sand or ash, created mortar. Vitruvius prescribed ideal ratios (De arch. 2.5.1). Volcanic ash from Puteoli near Naples and later similar ash from areas in the immediate vicinity of Rome contained both silica and alumina that reacted with the slaked lime to create a mortar that could also harden underwater. By the Imperial period Roman builders took advantage of local sources of this “pozzolana” sand (Lat: harena fossicia) that resulted in the characteristic reddish and black-grained mortars seen in so many buildings of Imperial date in the capital. The harbour project at the colony site of Cosa presents an early use (once thought to be 2nd century BCE but recently proposed as Augustan) of this technology. Builders even imported pozzolana from Italy to projects in the eastern Mediterranean, such as the estimated twenty-four thousand cubic metres used for the early Imperial harbour breakwaters at Caesarea Maritima in Judea.9
The mixing of caementa (fragments of rubble) with the mortar created the opus caementicium that forms the core of so many Roman foundations, walls, and vaults. The make-up of the caementa depended on the application and the materials available. Builders utilized stone fragments of all types—in Rome often chunks of tuff—and also rubble from demolished buildings, including pieces of broken brick, tile, and flooring. Foundations incorporated heavier materials, including basalt, while the cores of upper walls and vaults often contained vesicular volcanic scoria and pumice that reduced the weight of the superstructures.
To build a wall or a vault, labourers combined the caementa and the mortar layer by layer. For foundations on land, timber shuttering, the earthen walls of the foundation trench, or a retaining wall of fired brick, served as forms for the concrete. Concrete piers for bridges and breakwaters rose from wooden forms sunk in place to receive the uncured concrete. The hull of an enormous ship that had carried an obelisk from Egypt during the reign of Caligula famously served as the formwork for the outer breakwater of Claudius’s port project at the mouth of the Tiber (Suet. Claud. 20.3).
On walls, a durable facing of stone or brick concealed the opus caementicium core. Wall facings served as both forms during construction and as permanent outer sheathings. Archaeologists have classified these facings by appearance: opus incertum appears as a random matrix of stones for planar surfaces and small squared stones or fired brick as quoins for corners (fig. 11).
Opus quasi-reticulatum and opus reticulatum present facings in a near-diamond or diamond pattern respectively (fig. 12).
From the late 1st century bce and throughout the Imperial period, brick facings of brick (opus latericium, also called opus testaceum), or brick in combination with opus reticulatum (opus mixtum), were increasingly common (fig. 13).
Unlike their modern counterparts, the Roman fired brick is often triangular in shape, with one side forming the face of the wall and the two others facing the core, offering an ideal bonding face with the concrete. The facings as described in the order above developed in a more or less diachronic fashion from the mid-2nd century bce onwards, with reticulate work appearing in the latter half of the 1st century bce. Variations in the physical appearance and make-up of Imperial-period opus latericium also offer some clues as to dating; most useful is the analysis of brick stamps. In the late Roman period (4th century ce), as supplies of brick declined, alternating rows of brick and stone masonry were favoured (opus vittatum or listatum; examples of this method, however, are also found in earlier construction).
A final protective, often decorative, sheathing covered most wall facings, however attractive the underlying stone or brickwork. A combination of crushed brick or tile and mortar (cocciopesto) acted as a durable waterproofing agent, both to keep water in (as in cistern construction) and to keep it out (on the exterior exposed surfaces of vaults). Rough and fine plasters covered the majority of vertical exposed surfaces (opus albarium). Specialists (tectorii) mixed slaked lime and fine sand, the same ingredients used for mortars. The finest interior stucco work included a topcoat of lime gypsum and marble dust; these could take a high polish, even after painting.
Thin sheets of marble or other exotic stone covered the walls of the high-status public buildings and private residences. A layer of mortar and shims of marble scraps or terracotta provided the adhesive backing. Iron pins anchored to the wall and affixed to the edges of the marble held the slabs in place during construction.
The process of cutting and assembling patterned wall revetments of stone (opus sectile) was essentially the same in technique and appearance as that used for floor covering (fig. 14).
In terms of architectural scale and speed of construction, the adoption of opus caementicium was a game-changer. Cut-stone construction and tight-fitting timber joinery require a highly skilled labour force. Much of the process of involving opus caementicium was repetitive and—other than conceiving the design itself, the centring work, and the decorative finishes—involved much unskilled labour. The technology was well suited to a broad spectrum of public and private buildings, from the sprawling Imperial bath complexes to the multi-storied apartment houses of high Imperial Rome. Nevertheless, in those regions of the empire where cut-stone construction had enjoyed a long tradition and fine building stone was abundant—particularly in the eastern provinces—opus caementicium did not eclipse long-established practice.
Metals and Fastening Systems
Metals served as significant and in some applications essential components of Roman architecture. Blacksmiths shaped iron for use as clamps, tie bars, lewises (for lifting), nails, window frames, sheathing for piles, and all manner of straps and braces. Bronze, too, appeared in the form of pins, hardware such as hinges, sheathing for roofs, and even structural beams. Lead played another important role: sheets of lead sheathed roofs and could protect walls from water damage; molten led was used as a metallic adhesive in concert with clamps; and, of course, the plumbing of Roman buildings was often of lead. Although Romans were aware of the principle of the screw, metal screws were hardly, if ever, used as fasteners in Roman buildings.
Travertine blocks from the Metellan rebuilding of the Temple of Castor in Rome exhibit the earliest use of iron clamps in the capital (117 bce).10 Wood is also known to have been used for dovetail-shaped clamps, especially with tuff ashlar masonry, the Tabularium (78–65 bce), and the precinct wall of the Forum of Augustus (dedicated 2 bce) provide examples. Presumably the main advantage of wooden clamps was to prevent the ashlars from shifting during construction. H. Bauer suggested the use of iron tie bars for the vaulted portico along the forum facade of the Basilica Aemilia and in the Horrea Agrippiana, both built in the last decades of the 1st century bce.11 Later Imperial projects that employed vaulted porticoes, including the Basilica Ulpia in the Forum of Trajan (ded. 113 ce) and the Baths of Trajan, of Caracalla, and of Diocletian, all show evidence of the use of tie bars. Reconstructions show placement both below the vault (and thus visible) and embedded within the top of the vault.
While bronze is not as strong as iron under tension, it nevertheless offers some distinct advantages: it is attractive and durable, especially when employed in the form of relatively thin sheets. Thus bronze was prized as sheathing for buildings (especially roofing; e.g., the gilt-bronze tiles of the Temple of Jupiter Optimus Maximus; Plin. HN 33.57) or for elements such as doors, which could employ a laminate of bronze over a wooden frame (Plin. HN 34.13). In at least one instance, beams of bronze were used to frame the superstructure of a roof: the trussing of the porch of the Pantheon in Rome, now lost, was composed of bronze beams, U-shaped in section, interconnected with some form of rivet; a massive bronze pin from the structure still survives. Bronze column capitals evoked elegance and generous patronage both in Rome (Pantheon of Agrippa, late 1st century bce; Plin. Id.) and the provinces (the propylon of the sanctuary of Jupiter Heliopolitanus at Baalbek, 3rd century ce). Literary evidence indicates that a late reconstruction of Rome’s famous Temple of Janus was entirely of bronze (Dio Cass. 74.13.3). If extant parallels are any guide, one imagines a sheathing of bronze panels attached to a framework of wooden beams.
Thermodynamic analysis of large Imperial bathing structures indicates that large windows were in all probability glazed (see glass, Roman). Panes of glass must have been small and held in place within larger frames of iron. While floor mosaics (opus musivum, tessellatum, or vermiculatum) consist largely of cut-stone tesserae, glass mosaics offered brilliant colour palettes and pleasing light effects. Glass mosaic was used sparingly at first for decorating walls and ceilings, especially fountains and niches within walls and on the intradoses of vaults. It gained in popularity over time and by late antiquity played a significant role in the decoration of upper walls and vaults, especially those of early Christian tombs and churches.
Vaulting in Stone, Brick, and Concrete
The first experiments with vaulting in Rome may date back to the Archaic period (6th century bce).12 In these early applications, voussoirs of soft tuffs formed arches and barrel vaults suitable for utilitarian subterranean spaces like cisterns. City gates (e.g., the Porta di Giove at Falerii Novi, after 241 bce) represent the earliest extant large-scale surviving voussoir arches of tuff (fig. 15).
By the 1st century bce, travertine keystones and springers offered extra strength at points of stress (e.g., for the interior vaults of the Theatre of Marcellus). Limestone, travertine, and marble arches of cut-stone voussoirs, extremely durable and evoking tradition and prestige, remained an important feature of Roman architecture well after the introduction of concrete (fig. 16).
In the eastern provinces, where a tradition of stone masonry, fine materials, and a supply of trained craftsmen endured, barrel and cross vaults, domes, and half-domes testify to the possibilities of the finest cut-stone work.
In the west, the medium of concrete, sometimes in concert with stone or brick ribbing, was the preferred medium for the construction of volumetric vaulting. Barrel vaults in concrete appear by the mid-2nd century bce; by the end of that century, builders were experimenting with the highly novel forms then possible with the fluid medium of concrete. Inventiveness did not always radiate from the capital. The annular and coffered vaults of the hemicycles of the Fortuna Primigenia sanctuary in Tivoli (late 2nd century bce), for example, anticipate by 150 years the revolution in vaulted designs attributed to Nero in Rome. The earliest surviving concrete domes (first half of the 1st century bce) appear in the bath buildings of area around the Bay of Naples. These first successful attempts anticipate the sophisticated centrally planned designs with “pumpkin” or parasol domes that appear in the late 1st century (e.g., Domitian’s palace in Rome, late 1st century ce) and beyond.
Masonry vaulting presents two particular challenges: overall weight and problems associated with lateral thrust. Study of extant vaults indicates several ways designers managed to lessen overall weight. Coffering and variable thickness of the vault (a thinner cross section near the crown) reduced the overall mass. Lightweight caementa helped to decrease the density of the fabric of the vault. Another method involved the insertion of hollow amphorae into the concrete to mitigate the overall weight.
Buttressing countered lateral thrust. A side-by-side series of arches or barrel vaults is self-buttressing except for the terminal elements; at these points heavy piers were necessary. Embedding the springing of a dome deep within its cylindrical supporting wall (e.g., the Pantheon, early 2nd century ce) made free-standing rotundas possible. Barrel vaults placed at right angles to the bays of large cross vaults—such as those of the Imperial bath buildings and the Basilica Nova in Rome (early 4th century ce)—opposed lateral forces. At the tops of the piers of a cross-vaulted system, arched buttresses anticipated the medieval “flying” buttress solution (e.g., Markets of Trajan, 113 ce; Basilica Nova, early 4th century ce). “Step-rings” in masonry built along the shoulders of domes retarded the tendency for the curing or shrinking/expanding concrete to crack. Iron tie bars also worked from under and within the span of the vault to counteract lateral thrust (see “Metals and Fastening Systems” section).
Discussion of the Literature
Analysis of Roman building technologies has been a topic of interest since the Renaissance. The emergence of the field of Classical archaeology in the 19th century fostered the first modern systematic studies, such as Auguste Choisy’s L’Art de bâtir chez les Romans published in 1873. The topic continued to attract much scholarly attention in the opening decades of the 20th century. Women classicists played prominent roles as pioneers of this field, notably Esther Boise Van Deman (1862–1937) and Marion Elizabeth Blake (1892–1961). The former was the first to attempt a methodical system of dating based upon uses of Roman concrete and brick. Van Deman continued to focus on the materials of Roman construction until her death; Blake completed the last of Van Deman’s work yet died before her own study of construction was finished (Roman Construction in Italy from Nerva through the Antonines).13 Italian scholars continued with important contributions in the middle decades of the century. The Italian engineer and architect Gustavo Giovannoni published La tecnica della costruzione presso i romani in 1925. The ambitious excavation program at Ostia resulted in additional analysis of Roman construction methods that has continued to the early decades of the 21st century. Analysis of Roman building techniques was an important method for determining the phases of the city’s construction, as published in the first volume of the Scavi di Ostia series.14 The project in Ostia benefitted greatly from the participation of Herbert Bloch, who changed the study of Roman building practice and the dating of brick construction by his methodical studies of Roman brick stamps, work that first appeared in three substantial articles published between 1936 and 1938; an integrated version was published in 1947 and has been reprinted since.15 Studies of brick stamps and the Roman brick industry continue in the 21st century. Giuseppe Lugli, professor of ancient Roman topography at the La Sapienza for nearly three decades in the mid-20th century, published a two-volume set on Roman building technology in 1957, a contribution still in use.16
Jean Pierre Adam’s encyclopedic La construction romaine: Matériaux et techniques, first published in 1984 and translated into English ten years later, offers a comprehensive and richly illustrated digest of Roman building materials and techniques. Recent contributions by scholars like Rabun Taylor17 and Janet DeLaine18 have focused on process and logistics. The opening of the 21st century has seen new work on specific applications, for example, methods of vaulting by Lynne Lancaster19 and timbered construction by Roger Ulrich.20
Pioneering 19th-century and subsequent attempts to classify and date Roman building types by construction method have been questioned in recent years. The most prominent example is at Pompeii, where the division of the city’s building periods by material (e.g., the “limestone” or “tufo” periods) and their associated dates has been under revision as excavators have explored the levels below the destruction level of 79 ce. This new work has resulted in proposals for revising the dates of the introduction of concrete to central Italy and examining more closely the degree to which Rome influenced or was in turn influenced by new building technologies.21
- Pliny, Naturalis Historiae (Natural History). Completed 77 ce. Latin with facing translation by H. Rackam in the Loeb Classical Library, Cambridge, MA: Harvard University Press. 11 volumes. 1938–.
- Vitruvius, De architectura (Ten Books of Architecture). Late first century bce. Translation by Ingrid D. Rowland with commentary and illustrations by Thomas Noble Howe. Cambridge, U.K.: Cambridge University Press, 1999.
- Adam, Jean Pierre. Roman Building: Materials and Techniques. (English Translation of La construction romaine: Matériaux et techniques 1984). London and Bloomington: Indiana University Press, 1994.
- Blake, Marion Elizabeth. Ancient Roman Construction in Italy from the Prehistoric Period to Augustus. Washington, DC: Carnegie Institution, 1947; Roman Construction in Italy from Tiberius through the Flavians. Washington, DC: Carnegie Institution, 1959.
- Blake, Marion Elizabeth, and Doris Taylor Bishop. Roman Construction in Italy From Nerva through the Antonines. Philadelphia: American Philosophical Society, 1973.
- Bloch, Herbert. I bolli laterizi e la storia edilizia romana: contributi all’archeologia e alla stora romana. Rome: L’Erma di Bretschneider, 1968.
- Calza, Guido, ed. Scavi di Ostia. Topografia Generale. Rome: La Libreria dello Stato, 1953.
- Cifani, Gabriele. Architettura romana arcaica: edlizia e società tra monarchia e repubblica. Rome: L’Erma di Bretschneider, 2008.
- DeLaine, Janet. The Baths of Caracalla: A Study in the Design, Construction, and Economics of Large-Scale Building Projects in Imperial Rome. Portsmouth: Journal of Roman Archaeology Suppl. 25, 1997.
- Heres, Theodora. Th. L. Paries. A Proposal for a Dating System of Late-Antique Masonry Structures in Rome and Ostia (A.D. 235-600). Amsterdam: Rodopi, 1982.
- Hopkins, John North. The Genesis of Roman Architecture. New Haven, CT: Yale University Press, 2016.
- Lancaster, Lynne. Concrete Vaulted Construction in Imperial Rome: Innovations in Context. Cambridge, U.K.: Cambridge University Press, 2005.
- Lancaster, Lynne. “Roman engineering and Construction.” In The Oxford Handbook of Engineering and Technology in the Classical World. Edited by John Peter Oleson. Oxford: Oxford University Press, 2008.
- Lancaster, Lynne. Innovative Vaulting in the Architecture of the Roman Empire. 1st to 4th Centuries CE. Cambridge, U.K., and New York: Cambridge University Press, 2015.
- Lancaster, Lynne, and Roger Ulrich. “Materials and Techniques.” In A Companion to Roman Architecture. Edited by Roger Ulrich and Caroline Quenemoen. West Sussex: Wiley Blackwell, 2014.
- Lugli, Giuseppe. La tecnica edilizia romana: con particolare riguardo a Roma e Lazio. 2 vols. Rome: Eredi Dott. G. Bardi, 1957.
- Malacrino, Carmelo. Constructing the Ancient World. Architectural Techniques of the Greeks and Romans. Los Angeles: J. Paul Getty Museum, 2010.
- Taylor, Rabun. Roman Builders: A Study in Architectural Process. Cambridge, U.K., and New York: Cambridge University Press, 2003.
- Ulrich, Roger. Roman Woodworking. New Haven, CT: Yale University Press, 2007.
- Vitti, Paolo. Building Roman Greece: Innovation in Vaulted Construction in the Peloponnese. Rome: L’Erma di Bretschneider, 2016.
- Ward-Perkins, John Bryan, and Hazel Dodge. Marble in Antiquity: Collected Papers of J. B. Ward-Perkins. London: Archaeological Monographs of the British School at Rome, 1992.
1. Marcello Mogetta, “A New Date for Concrete in Rome,” Journal of Roman Studies 105 (2015): 1–40.
2. John North Hopkins, The Genesis of Roman Architecture (New Haven, CT: Yale University Press, 2016).
3. Herbert Bloch, I bolli laterizi e la storia edilizia romana: contributi all’archeologia e alla stora romana (Rome: L’Erma di Bretschneider, 1968).
4. Lynne Lancaster, Concrete Vaulted Construction in Imperial Rome: Innovations in Context (Cambridge, U.K.: Cambridge University Press, 2005), 88.
5. Lynne Lancaster, Innovative Vaulting in the Architecture of the Roman Empire (Cambridge, U.K., and New York: Cambridge University Press, 2015), 99–128.
6. Lynne Lancaster, “Roman Engineering and Construction,” in The Oxford Handbook of Engineering and Technology in the Classical World, ed. John Peter Oleson (Oxford: Oxford University Press, 2008), 275.
7. Marcello Spanu, “L’impiego di anfore nelle volte romane e tardo-antiche: distribuzione e modalità,” Daidalos: Studi e ricerche del Dipartimento di Scienze del Mondo Antico 8 (2007): 185–223.
8. Filippo Coarelli, “Public Building in Rome Between the Second Punic War and Sulla,” The Papers of the British School at Rome 45 (1977): 1–23: beginning of the 2nd century bce or slightly earlier; cf. Marcello Mogetta, “A New Date for Concrete in Rome,” Journal of Roman Studies 105 (2015): 2–7: domestic use in the mid-2nd century.
9. Robert Hohlfelder, “Constructing the Harbour of Caesarea Palaestina, Israel: New Evidence from ROMACONS Field Campaign of October 2005,” International Journal of Nautical Archaeology 36 (2007): 409–415; and, for chronological development, Christopher J. Brandon, Robert L. Hohlfelder, Marie D. Jackson, John P. Oleson, Building for Eternity: The History and Technology of Roman Concrete Engineering in the Sea (Oxford: Oxbow Books, 2014).
10. Lynne Lancaster, Concrete Vaulted Construction in Imperial Rome: Innovations in Context (Cambridge, U.K.: Cambridge University Press, 2005): 113.
11. Heinrich Bauer, “Un tentativo di ricostruzione degli Horrea Agrippiana,” Archaeologia Classica 30 (1978): 132–146; Heinrich Bauer and A. Pronti, “Elementi architettonici degli Horrea Agrippiana,” Archaeologia Classica 30 (1978): 107–131; and Heinrich Bauer and A. Pronti, “Basilica Aemilia,” in Kaiser Augustus und die verlorene Republic: eine Austellung im Martin-Gropius-Bau, Berlin, 7. Juni—14 August 1988, eds. Hofter et al. (Mainz: Philipp von Zabern, 1988), 200–212.
12. Gabriele Cifani, Architettura romana arcaica: edilizia e società tra monarchia e repubblica (Rome: L’Erma di Bretschneider, 2008).
13. Marion Elizabeth Blake and Doris Taylor Bishop, Roman Construction in Italy from Nerva through the Antonines (Philadelphia: American Philosophical Society, 1973).
14. Guido Calza, ed., Scavi di Ostia. Topografia Generale (Rome: La Libreria dello Stato, 1953).
15. Herbert Block 1936, 1938, 1947.
16. Giuseppe Lugli, La tecnica edilizia romana: con particolare riguardo a Roma e Lazio, 2 vols. (Rome: Eredi Dott. G. Bardi, 1957).
17. Rabun Taylor, Roman Builders: A Study in Architectural Process (Cambridge, U.K., and New York: Cambridge University Press, 2003).
18. Janet DeLaine, The Baths of Caracalla: A Study in the Design, Construction, and Economics of Large-Scale Building Projects in Imperial Rome (Portsmouth: Journal of Roman Archaeology Suppl. 25, 1997).
19. Lynne Lancaster, Concrete Vaulted Construction in Imperial Rome: Innovations in Context (Cambridge, U.K.: Cambridge University Press, 2005); Lynne Lancaster, Innovative Vaulting in the Architecture of the Roman Empire. 1st to 4th Centuries CE (Cambridge, U.K., and New York: Cambridge University Press, 2015).
20. Roger Ulrich, Roman Woodworking (New Haven, CT: Yale University Press, 2007).
21. Michael Fulford and Andrew Wallace-Hadrill, “Towards a History of Pre-Roman Pompeii. Excavations beneath the House of Amarantus (I 9, 11–12),” The Papers of the British School at Rome 67 (1999): 37–39; and Marcello Mogetta, “The Early Development of Concrete in the Domestic Architecture of Pompeii,” Journal of Roman Archaeology 29 (2016): 43–72.